emerging-channels
Pass
Audited by Gen Agent Trust Hub on May 20, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions in
SKILL.mdinclude a step to execute a local Python script,campaign-tracker.py, to list campaign history. This script is an environmental dependency intended for internal workflow management and is not provided within the skill package. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest brand-specific data and guidelines from the
~/.claude-marketing/directory to customize output. This creates an ingestion surface where external content could influence the agent's behavior. - Ingestion points: Brand profile and guideline files (JSON/Markdown) located in the application's configuration directory (e.g.,
profile.json,restrictions.md). - Boundary markers: Absent; the agent is instructed to load and apply these rules directly into its operating context.
- Capability inventory: Local command execution via Python for tracking campaigns.
- Sanitization: The skill does not specify validation or sanitization for the content of the ingested files.
Audit Metadata